[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: Stata (gllamm) benchmarks for different platforms?
> The (amateur) benchmarks I have seen (for an earlier version of XL) are very
> impressive and suggests that the gcc and the XL doesn't even play in the
> same league. If you are really serious about producing high performance
> code, the IBM XL compiler seems to be the obvious choice. And this works for
> other Macs as well (at least the G4 and the G5).
I tried it months ago and had to pass on it. Yes, it produces fast code...
if you use libraries and compiler flags that trade precision for speed which
obviously we're not willing to do. The fast code it produced didn't even come
close to passing our certification tests nor do they claim that it would
(gcc 3.3 has similar flags yet the documentation even says that it's unsafe
for some situations).
> Unfortunately, you are probably right. I still though, have a hard time
> accepting the G5 to have only about 27% of the Opteron performance. If so,
> the guys at Virginia tech really made a poor choice building a supercomputer
> on the G5 instead of the Opteron. And they still managed to get to third
> place in the world.
It's performance is not 27% of the Opteron... one or two programs do not
define a processor's performance. There are situations where the G5 can be
slower just as there are situations where it can be much faster. That's why
benchmarks contain a whole suite of tests and include real world tests with
applications rather than just artificial programs. Just because gllamm does
not perform as well on the G5 does not mean the G5 doesn't outperform other
processors in other areas.
* For searches and help try: